Case 2

Capital Budgeting with

Staged Entry

Maastricht University
Faculty of Economics and Business Administration
Maastricht, November 16 2004
Arends, Tangela I199524
Middelbeek, Robbert I168165
Pollaert, Rian I161446
Group 2
Subgroup 2
Tutor: Nils Kok
Financial Management & Policy
Report Case 2

0
In the seafood industry, Gulf Coast Fisheries is the leading seafood harvester and
processor. At the moment, they concentrate only on a variety of saltwater seafood,
however, due to increased competition they are considering to move part of their
operations into the freshwater catfish market. To examine if this shift is profitable and
which strategy they should adopt, we will come up with a recommendation based on
the following questions.

1.
Consider the land acquisition.
a) What cost, if any, should be attributed to the catfish project?
The firms already owns a suitable tract of land that can be sold now for $ 1.500.000,
but other suitable sites can also be purchased for this amount. Therefore, the costs that
should be attributed to the catfish project amounts $ 1.500.000. Gulf Coast Fisheries,
INC. has the option to acquire additional land for the Shrimp Division; this has not to
be taken in account considering the cost of land acquisition. If the land is used for the
catfish project, the Shrimp Division should purchase a new site by exercising the
option. The option that will pay $ 100.000 in December 1995 and $ 1.900.000 in
December 1999, is more valuable than buying land today for a price of $ 1.500.000.
The following question will give a more precise answer to this.

b) Assuming that the currently owned site is used for this project, how should the Gulf Coast
Shrimp Division obtain a site? What discount rate should be used in analyzing the option
alternative?
Using an annual appreciation rate of 9 percent, the value of the site can be calculated
as follows: $ 1.500.000 * 1,094 = $ 2.117.372,4. Next, this value can be used to
calculate the option gain: $ 2.117.372,4 - $ 1.900.000 / (1,04) 4 = $ 185.810,8. The
general inflation rate is 4 percent. Finally, the profit or loss of exercising the option
can be computed: $ 185.810,8 - $ 100.000 = $ 85.810,8. The option price is subtracted
of the option gain, that gives a option profit of $85.810,8.
In stead of buying the land today for $ 1.500.000, the Shrimp Divisions can earn a
profit. The discount rate used in the calculations is the inflation rate of 4 percent,
which is assumed to be the risk free rate.

1
2.
a) What would be the R&D cash flows in 1997 through 2003? Should any R&D cash flow for
1995 be included in the analysis?
In total, $ 1.000.000 was spend on R&D in 1995, of which $ 400.000 has been
expensed for tax purposes. With tax consideration, the amount of $ 240.000 ([1-
0.4]*400.000) is the cash outflow. The remaining $ 600.000 will be amortized over
the first 3 years of the operating life of the new facility. Cash flow is $80.000
(600.000*0.4/3).

b) How is salvage value taxed?

The salvage value is the market value of an asset after its useful life. Both buildings
fall into the MACRS 31.5-year class, and will be depreciated starting in 1997. The
building considered under Plan L, has a salvage value of $ 3.880.000 (5000-
[5000*0.032*7]) after seven years. It is stated that the buildings could be sold for
about half their book value at the end of 2003. This is then $ 1.940.000. The taxation
of the salvage values depends on prescribed tax regulation. Therefore, it is hard to
determine taxes to be paid.

c) Complete the large plant high demand high growth cash flow statement.
The large plant high demand high growth cash flow statement in concluded in the
Appendix.

3. Assume that the large plant project is judged to be of average risk. What are its stand-
alone NPV, IRR, MIRR, and payback if initial demand and growth are both high? What are
the values for these same variables if the small plant is built and demand growth are again
high?
For the large plant project with high demand and high growth the following numbers are calculated:

NPV 19,867
IRR 23.31%
MIRR 18.46%
Payback 7.055

The same calculations for the small plant project with high demand and high growth are given in the following table:

NPV 10,049
IRR 18.86%
MIRR 12.66%
Payback 7.24

4.
a) Examine the decision tree for the large plant, and discuss what it is designed to do and
how it works. Focus first on the left section, where the probabilities are given; then describe

2
the cash flows; then explain the NPVs; then explain what the joint probabilities are, what they
mean and how they were calculated; then explain what the products are; and finally explain
the expected NPV, standard deviation, and the coefficient of variation terms.
The left section of the decision tree first gives the probabilities of high and low
demand. Second, the percent change of growth and the corresponding probability is
given, either high or low. In total, there are four possible outcomes, high demand with
high growth etc. Third, the Net Present Values are given, calculated for each of the
four outcomes and discounted at a discount rate of 10 percent. The joint probability
indicates the probability of for example, high demand and low growth. Next, the NPV
for each case is multiplied by the joint probability of that case. This calculation gives
the expected NPV for the entire project. The expected NPV shows whether the project
was profitable. The standard deviation is the volatility of the project, the higher the
standard deviation the more volatile the NPV of the project. The coefficient of
variation shows the risk per unit of returns and is calculated as the ratio of standard
deviation over the expected NPV.

b) Explain the abandonment lines and discuss why the product terms are based on
abandonment.
An abandonment option exists if the company decides to give up the project at an
early stage. The salvage value at the moment of possible abandonment at the end of
1999 is calculated. The salvage value is the initial investment minus the cumulative
depreciation until 1999. This would create new NPV’s for each abandonment option.
Then the option to abandon or not that gives the best result was used in the calculation
of the expected NPV.

c) If the contract with the city of Pascagoula would rule out the possibility of abandonment
prior to 2003, what would this do to the project’s expected profitability and risk?
First, the project’s expected risk would increase. Second, the expected profitability
would decrease.

5.
a) Now consider the small plant decision tree. Why are there more branches on that tree than
on the large plant tree?

Considering the small plant project, there are more branches because there is the
opportunity of expansion, which is not the case for the large plant project.

3
b) Why does each branch not have a joint probability and product? Why are those terms
shown for the branches where they appear?
Each branch does not have a joint probability and product because the four
probabilities that were also stated before are the only options the company is
interested in. Other options would give a lower profit or higher loss. Expansion is
considered solely in the scenario of high demand and high profit.

6.
Based on the data in the decision tree, which plant has the higher expected NPV? Which
appears to be riskier? Do you think the same cost of capital used to evaluate both plants?
The expected NPV for the small plant is positive and thereby higher than the one of
the large plant, which is negative. The riskiness of the large plant is higher, which can
be seen by looking at its coefficient of variation. Indeed, the large plant has in
absolute value a coefficient of 1,75 while the small plant has only a coefficient of
0,15, which is less risky. This seems to be logical since one can decide whether to
expand or not in the small plant. Following from this fact, the large plant project
should be discounted at a higher cost of capital.

7.
If most of Gulf Coast’s projects have a variation coefficient in the range of 0.5 to 0.7, how
might this information be used in the analyses and what effect might it have?
The variation coefficient for the large plant project is 1.75 and for the small plant
project it is 0.15. The latter is lower than the average variations coefficients of Gulf
Coast’s projects, but the first one is higher. Concluding from this finding, the large
plant project is notably riskier than the small plant project. The effect they have is that
the large plan should be discounted at higher cost of capital than the corporate rate,
similarly the small plan should be discounted at a lower rate.

8.
Suppose you were told that your boss has reassessed the probabilities. How should the
situation change if these new probabilities were assumed? (HD = 0.9; HG = 0.9 or HD =
0.6; HG = 0.5)

4
Which set of revised probabilities would favor the large plant? Which would improve the
relative status of the small plant?
Starting with the first change (0.9, 0.9) the following changes occur:
Large plant changes from

EXP(NPV) -9216,34
SD (NPV) 16138,15
Coefficient of variation -1,75104
To:

EXP(NPV) -10507,6
SD (NPV) 24293,94
Coefficient of variation -2,31204
Small plant changes from

EXP(NPV) 16060,22
SD 2488,609
Coefficient of variation 0,154955
To:

EXP(NPV) 18378,03
SD 3739,717
Coefficient of variation 0,203488

And with the second change (0.6, 0.5) the following changes occur:
Large plant changes into:

EXP(NPV) -9221,07
SD (NPV) 10268,01
Coefficient of variation -1,11354

Small plant changes into:

EXP(NPV) 13297,76
SD 1409,243
Coefficient of variation 0,105976

It can thus be seen that in the first scenario of the probabilities (0,9;0,9) the NPV of
the large plant gets worse, the standard deviation increases, and the coefficient of
variation (in absolute value) rises accordingly. For the small plant the NPV rises, but
the standard deviation rises relatively more which increases the coefficient of
variation.
In the second scenario with the probabilities (0.6;0.5), the NPV of the large plant
remains about the same, but the standard deviation decreased a lot, which in turn
lowers the riskiness of the project. For the small plant, the standard deviation declines
relatively more than the expected NPV so the coefficient of variation decreases.

5
9.
According to Brigham and Daves, the Sensitivity analysis is a technique which
measures the change in NPV, as a response to changes in input variables such as price
or sales quantities. Although, in this analysis only one variable is changed at the time.
In the sensitivity analysis there is always the base-case situation, which is the situation
that uses the most probable inputs. This base-case situation will provide the base-case
NPV, and then the sensitivity of the change of 1 input will be measured. I.e. what will
happen to the base-case NPV if the sales price rises with 10%? The more sensitive the
NPV is to the sales price, the higher the percentage-deviation will be from the base-
case NPV. In the case of Gulf coast, some inputs do not have to be changed, i.e. the
construction costs of the 2 plans are merely fixed, and so are the capacities of the
plants. However, the variable costs, projected at respectively 60% and 65% -of the
sales- for the large and the small plant can vary. The fixed costs, that exclude
depreciation, managerial salaries and taxes, are also an estimation as a function of
future inflation. And even more uncertain are the sales-price estimates, sales
quantities and increases in sales rates. All these factors are most important in
conducting a sensitivity analysis, and one at the time should be changed to measure
their effect on base-case NPV. In the table displayed below, results of a small
sensitivity analysis can be found, in which the small-plant case is used, assuming no
abandonment before 2003, and a high-growth, high demand situation. The variable
costs were changed with 10% in each direction, leading to an enormous change in
NPV (123% in either direction), implying a huge influence of variable costs on the
Net Present value. The 10% change of fixed costs leads to a more reasonable,
although large change in NPV. The change in price-growth and demand growth rates
by 1% in either direction leads to a comparable high influence on the NPV, which
might imply that in this case we are dealing with a high risk project. If the estimates
deviate largely, it may have large consequences for the project. As described in the
case, the WACC will depend on the risky ness of the project, but these numbers will
already give an indication that we are dealing with a higher risk project.

6
 Small plant, base-case: High growth, high demand
NPV in basis situation: 10.049
Change in variable New NPV change
+10% in variable costs -2328 -123,2%
-10% in variable costs 22426 123,2%
+10% in fixed costs -6693 -17,3%
-10% in fixed costs 20767 17,3%
1% additional price growth 12699 26,4%
Also in 1% less price growth 7512 -25,2% the
1% additional demand growth 12479 24,2%
Large 1% less demand growth 7721 -23,2% firm,
high-growth high-demand case the influence of the variable costs are largest, although
the effect is much lower than in the small-plant case, implying a possible lower risk
for this project.

Large plant, base-case: High growth, high demand

NPV in basis situation: 19867
Change in variable New NPV change
+10% in variable costs 8071 -59,37%
-10% in variable costs 31662 59,37%
+10% in fixed costs 16395 -17,47%
-10% in fixed costs 23338 17,47%
1% additional price growth 22910 15,32%
1% less price growth 16946 -14,70%
1% additional demand growth 23090 16,22%
1% less demand growth 16781 -15,53%

10.
The scenario analysis is a little more advanced than the sensitivity analysis, because in
this analysis not only 1 variable at the time is changed, but total ‘scenarios’ are
developed. Variables that tend –or are probable- to move together can be included in
the scenario. I.e., during economic downturn, not only the sales quantity might
decline, but also the sales price. Usually, a worst-case scenario and a best-case
scenario are set up next to the base-case, each having their own probability to occur.
In the case of Gulf coast, the best case scenario would be the high-growth high-
demand case (probability: 0,8*0,7 = 0,56), and the worst case scenario these two
factors both would be low (probability 0,2*0,7 = 0,14). The probabilities of the
scenarios can be deduced from the decision tree, and in our case we can use as a base

7
case the accumulated probability of low-demand high-growth, and high-demand-low
growth (probability: 0,8*0,3 + 0,2*0,3 = 0,3).
11.
The Monte Carlo simulation is a sort of automated scenario and sensitivity analysis at
the same time. It randomly assigns different values to the variables, taking into
account the probabilities (can be found in decision tree) and standard deviations of
each variable. Also correlation between 2 variables is taken into account so variables
that are more likely to move together (thus, highly correlated), will be changed
accordingly. The simulator will repeat the process thousands of times, to develop a
probability distribution, from which i.e. standard deviations of the NPV can be
deduced.
12.
Looking at the decision trees and at the project NPV’s, one can see that in general the
Small plant will generate higher NPVs and also lower standard deviations in
comparison to the large plant. This is why we advice to invest in the small plant, even
though the sensitivity analysis showed a higher sensitivity of the NPV to variable
costs, fixed costs, price growth and demand growth. But, the sensitivity analysis does
not take into account any reasonable probabilities and correlations to asses the risk. A
Monte-Carlo simulation would have to be performed to get a more reliable assessment
of the risk. So, the final advice is to invest in the small plant.